US: The market for artificial intelligence (AI) products and services is expected to grow rapidly, potentially reaching up to US$990 billion (about S$1.27 trillion) by 2027. This projection comes from Bain & Company’s fifth annual Global Technology Report, released on Sept 25.
According to the report, the market for AI hardware and software could expand by 40% to 55% each year, reaching between US$780 billion and US$990 billion by 2027.
Nvidia’s CEO, Jensen Huang, pointed out during the company’s 3QFY2024 earnings call: “Generative AI is the largest total addressable market expansion of software and hardware seen in several decades.”
The Edge Singapore reported that the demand for AI workloads will increase by 25% to 35% annually until 2027. This surge is likely to drive significant growth in data centres, with their power needs growing from the current 50 to 200 megawatts to more than a gigawatt in the coming years.
The costs associated with this expansion could reach between US$10 billion and US$25 billion over five years, impacting areas like infrastructure engineering, power production, and cooling systems.
The demand for graphics processing units (GPUs) is also projected to rise by 30% or more by 2026. This demand may strain supply chains for chips used in data centres, personal computers, and smartphones. Bain also warns that ongoing global tensions could lead to another semiconductor shortage.
Bain said that the rise of sovereign AI adds another challenge for tech companies. However, it also opens up opportunities for governments globally, many of which are investing billions to support sovereign AI initiatives. These investments focus on building local computing infrastructure and developing AI models created and trained using domestic data.
As companies face challenges in managing supplies and protecting data, Bain noted a growing interest in small language models. These models use algorithms that assist with computing, networking, and storage tasks, becoming essential in the current landscape.
A survey by Bain of over 200 companies found that generative AI could save about 10% to 15% of software engineering time. This finding places pressure on software development firms to improve efficiency.
Roy Singh, Bain’s global head of advanced analytics, mentioned that if implemented correctly, generative AI could lead to efficiency gains of 30% or more. However, achieving this requires more than just introducing coding assistants into processes.
This surge in interest comes at a time when many software companies are seeing slower revenue growth. Bain’s analysis revealed that the median annual revenue growth for around 90 publicly traded software-as-a-service (SaaS) companies dropped by 16 percentage points in the last two years.
As a result, these companies have cut back on their sales and marketing spending. Sales and marketing budgets fell from 41% of revenue in 2022 to 33% in 2024, while spending on research and development dropped slightly from 21% to 18%.
Bain’s report also highlighted that mergers and acquisitions (M&A) in the tech sector are becoming less predictable. Regulatory challenges have led companies to focus on “scope” deals, which are aimed at acquiring new capabilities and products or entering new markets, rather than large-scale mergers.
Between 2015 and 2018, scope deals in the tech industry rose from 50% to 80% and have remained steady. Over the last six years, scope deals have made up nearly 80% of all M&A in the tech sector.
These deals largely depend on revenue synergies. According to Bain’s research, the biggest challenges in achieving these synergies are difficulties in integrating product portfolios (36%) and in executing successful go-to-market strategies and transformations (35%). /TISG
Read also: Tech Week Singapore 2024: AI spotlight with the world’s top tech leaders
Featured image by Depositphotos (for illustration purposes only)